em driver

Jack Vogel jfvogel at gmail.com
Tue May 15 04:50:48 UTC 2007


On 5/14/07, Randy Bush <randy at psg.com> wrote:
> LI Xin wrote:
> > Randy Bush wrote:
> >> could the em driver have been borked by the recent changes such that
> >> once a day or so it went out for a few minutes to an hour?  if so, did
> >> the changes three days ago fix it?
> > What do you mean by "went out"?  According to my understanding 1.175 of
> > if_em.c seems to fix the case where em(4) fails to attach on certain
> > hardware...
>
> running fine.  stops moving packets for five to 60 minutes.  starts moving
> packets again.  all with no intervention, in rack, long way away.  e.g. see
> two stops in
>
>     http://rip.psg.com/~randy/070514.em0-borkage.png
>
> no other hosts with same symptom, though three with em ethers updated at
> same time.
>
> and, of course, it used to work fine until cvsup of May 8 12:03 (gmt).
> everything always "used to work." :)

What do you expect anyone to do based on this data??? For all I
know someone in another room is turning off a switch every couple
hours :)

So if you want to report a problem, put yourself in the place of
the engineer who has to fix it, what would you want... well, how
about hardware description, pci config, log file output, I dont
know, something that would give me a chance of knowing what
the problem is.

Jack


More information about the freebsd-current mailing list